A headline in the Herald reads: "Farmers warned to prepare for the worst as risk of El Nino grows." The report states: "The National Institute of Water and Atmosphere has given a 50 per cent chance of an El Nino developing over winter."

Quality risk communication involves providing the public with information that is useful. A 50:50 prediction is not. It is no better than flipping a coin.

Worse, it can lead to an impression that skill exists when it does not.

For instance, send an email to, say, 512 people with the message that there is a 50 per cent chance of an El Nino developing over the coming summer. Of these, 256 should get the correct prediction. You email them another 50:50 climate prediction; this time 128 receive the correct prediction twice in a row, and so on until a useful number of people are convinced you have some skill at forecasting future climate.


To the climate scientist, a 50:50 prediction is a safe prediction; as safe as predicting, say, that next summer's rainfall will be about the long-term summer average. The probability of occurrence is usually always greater around the mean.

Climate systems are incredibly complex. In words attributed to Albert Einstein: "When the number of factors coming into play in a phenomenological complex is too large, the scientific method in most cases fails. One need only think of the weather, in which case the prediction even for a few days ahead is impossible."

The problem is complicated by the fact that the public usually fear the worst, and fear sells. So, if the period for which the prediction is made is beyond the end of the climate scientist's lifetime, such as with long-term predictions of human-caused climate change, or "global warming", any scary prediction will attract attention and hopefully also research funds or job promotion.

Many experts passionately declare they believe future climate will be dramatically different due to human action. However, the challenge in climate science is correctly attributing cause.

Getting back to El Nino, the phenomenon is in fact a potentially important indicator of short-term inter-annual and inter-seasonal rainfall in New Zealand. Many climate scientists consider an understanding of the development and course of El Nino to be a key to successful long-range climate forecasts. But to date this understanding is poor.

Among the most promising results are assessments using the so-called "phases" of El Nino development. Some research has shown that monthly phases are correlated with long-term rainfall records. The results are expressed in terms of probability distributions for the various climate zones around New Zealand. As El Nino phase developments relate to the actual rainfall amounts, expressed in terms of probability distributions, the phases have direct application to the agricultural and farming sectors of New Zealand. But the reliability of the results are mixed and not yet considered to be reliable.

An alternative strategy to prediction - or better a convoying strategy - is a focus on adaptation. We have the knowledge and skills to reduce the impacts of excessively wet or dry periods.

A key to reducing losses is to implement measures that increase community awareness, readiness and resilience. The emphasis needs to be on dealing with the social, political and economic impediments that prevent effective flood-risk or drought-risk reduction.


Chris de Freitas is an associate professor in the School of Environment at the University of Auckland.
Debate on this article is now closed.